Anthropic valuation surge and Zoom's $1B+ windfall

The Bloomberg-led reporting cycle hooked on the valuation, but the more under-discussed move is on Zoom's balance sheet. According to a regulatory filing made Friday, Zoom Ventures' approximately $51M investment in Anthropic in May 2023 is now valued at nearly $1.3B — roughly a 25x return on a stake that originated as a strategic partnership to bring Claude into Zoom's product stack ^[2]. That mark is anchored to Anthropic's February Series G at $380B post-money ^[5], not the in-flight $900B+ round.

Wedbush has modeled Zoom's stake at the $900B mark in a wide $4.11B to $11.31B range — depending on whether the original instrument was straight equity or carried any preference-stack adjustments ^[4]. CEO Eric Yuan framed the broader story around Zoom being adopted as an 'AI-first system of action for modern work' ^[6], but the market is reading something narrower: ZM is now one of the only liquid, public-equity proxies for Anthropic ahead of IPO. The stock jumped roughly 9% after the earnings beat and Anthropic windfall disclosure, and the price action reflects that re-rating as much as the SaaS fundamentals ^[6].

Step back from the absolute numbers and the more arresting fact is the slope. Anthropic raised its Series E at $61.5B post-money in March 2025 ^[7], its Series F at $183B in September 2025 ^[8], and is now closing a round above $900B ^[1]— a roughly 15x re-rating in 14 months. The February Series G alone landed at $380B, which felt aggressive at the time and now reads as a stepping stone ^[5].

What changed in that window is not vibes — it is revenue. Anthropic told investors it expects Q2 revenue of $10.9B, more than doubling quarter over quarter, with an annualized run rate expected to surpass $50B by the end of next month ^[1][9]. The earlier reporting from VentureBeat traced the run rate from roughly $87M in January 2024 to a $30B mark — an 80x trajectory that, paired with hyperscaler capacity commitments from Google and Amazon (up to $30B and $20B respectively in milestone funding), is what investors are actually underwriting ^[10][11]. Against a $50B forward run rate, $900B collapses to roughly an 18x multiple — rich, but not preposterous against software comps for a category leader.

On private secondary market Forge Global, Anthropic shares have already traded to imply a roughly one-trillion-dollar valuation, ahead of OpenAI's $880B mark in the same venue ^[3]. Saints Capital's Ken Sawyer flagged that an Anthropic shareholder recently offered to unload shares at a $1.15T valuation ^[3]. The gap between the primary round ($900B) and these secondary prints isn't a contradiction — it's a scarcity tax.

Primary rounds price the marginal share against committed capacity and contractual rights such as board seats, pro-rata, and preference stack. Secondaries price tiny slivers of stock that hedge funds and family offices are willing to overpay for because there is no other way to get exposure pre-IPO. With Anthropic's float on Forge thin and demand from institutional crossover funds intense, $1T-plus secondary marks should be read as a liquidity premium, not a fair-value signal. TechCrunch reported Anthropic asked investors to submit allocations within 48 hours — the kind of compressed process that only works when demand vastly exceeds supply ^[12].

Anthropic also disclosed it is on track for its first profitable quarter — and that's where the optimist/skeptic split sharpens ^[9]. The believer view, articulated by Coatue's Philippe Laffont, leans on agentic coding as a durable enterprise wedge; CFO Krishna Rao framed the round as a direct read on customer demand, and GIC's Choo Yong Cheen called Anthropic 'the clear category leader in enterprise AI' ^[5].

The pushback is more technical. Reading the operating-profit headline literally misses that training capex is the largest cash line at any frontier lab, and it doesn't sit inside operating income. Inference margins on frontier models are structurally tighter than pure-play SaaS, because every incremental token routes through GPU time that doesn't compress the way storage and bandwidth did. That gap is why the $50B run-rate target matters more than the profitability headline: it's the only number large enough to absorb both training capex and inference COGS while still leaving a margin profile that justifies a $900B+ mark ^[1].

OpenAI's most recent primary mark is $852B from a round completed in March 2026 ^[11]. Anthropic is now poised to clear it on both the primary ($900B+) and secondary ($1T vs $880B) markets ^[3]. The picture analysts are forming is one of a category leader pulling away on both pricing and traction.

The second-order effect is on capital allocation across the rest of the AI stack. Sequoia, Dragoneer, Altimeter and Greenoaks committing roughly $2B each into Anthropic is roughly $8B of growth-stage capital sucked out of the system that would otherwise have funded a long tail of AI infrastructure, agent, and applied-AI startups ^[1]. With Google and Amazon also pre-committed for up to $30B and $20B more in milestone-based deployments, the cap table is now densely populated with the two cloud providers and the four largest growth funds — a structure that will shape both Anthropic's IPO mechanics and the competitive dynamics for any rival lab still trying to raise at meaningfully smaller valuations ^[11].